1.6 Metric Baseline
Describe setting a metric baseline
A metric baseline is a reference point that represents the normal or expected performance of a network or system. It is established by collecting and analyzing performance data over a specific period. Baselines are essential for network monitoring and troubleshooting as they allow you to:
- Assess performance: Compare current measurements against the baseline to determine if the network is performing as expected.
- Detect anomalies: Quickly identify deviations from the baseline that may indicate a performance issue or security threat.
- Troubleshoot problems: Use the baseline to understand the magnitude of a problem and how it differs from normal behavior.
Key Concepts
Establishing a metric baseline involves several key steps:
-
Define metrics and objectives: Determine which metrics are crucial for monitoring network health, such as bandwidth utilization, CPU usage, or application response time. Clearly define what you aim to achieve by establishing a baseline for these metrics.
-
Choose a time period: Select a timeframe for data collection that captures representative network behavior, including both peak and off-peak traffic patterns. The length of this period depends on the stability and variability of the network environment.
-
Collect data: Gather relevant data points from appropriate sources, such as ThousandEyes tests, SNMP monitoring, or network device logs.
-
Analyze data: Examine the collected data to identify patterns, trends, and typical performance levels. Use statistical methods to determine average values, percentiles, or standard deviations for each metric.
-
Establish baseline: Based on the data analysis, set a reference point that represents the normal or expected performance for each metric.
-
Monitor and update: Continuously monitor network metrics over time and update the baseline as network conditions or requirements evolve.
-
Define alert thresholds: Set thresholds based on the baseline values to trigger alerts when metrics deviate significantly from the expected range. This allows you to proactively detect and address potential performance issues.
Resources
- Getting Started with Dashboards
- Creating and Editing Alert Rules
- Dynamic Baselines
- The Art of Designing ThousandEyes Alert Rules
- Create a Baseline Dashboard
- Event Detection
Sample Questions
1.6 Question 1
A network administrator wants to establish a baseline for CPU utilization on their core routers. Which data source would be MOST appropriate for this purpose?
- A) DNS resolution time from ThousandEyes tests
- B) HTTP Server response times from ThousandEyes tests
- C) SNMP data collected from the routers
- D) Network path visualization from ThousandEyes tests
1.6 Question 2
What is an important consideration when choosing a time period for collecting data to establish a baseline for interface utilization on a critical network link?
- A) Selecting the time period with the lowest network traffic volume.
- B) Ensuring the time period aligns with the organization's financial year.
- C) Capturing both peak and off-peak traffic patterns for a representative view.
- D) Limiting the time period to minimize the amount of data that needs to be analyzed.